How Just Move In split the work between AI and the team

Recently we spoke to Ross Nichols, co-founder of Just Move In, about why they rebuilt their 10-year-old platform around AI. This week we’re looking at how their AI, Jay, and the human team work together day to day.

Before Jay, a move specialist would spend 20 to 30 minutes on a single call. Every customer started from scratch. The specialist walked them through the full process – what services they needed, when to switch, what options were available. But that approach couldn’t scale – to offer services beyond energy, broadband and insurance, they’d have needed a much larger team.

Jay launched in mid-December. Energy is the first service running through the AI. Insurance and broadband follow by end of Q1, each added individually because every supplier journey has different nuance.

Personalisation starts with the first message

Just Move In has mapped over two dozen customer personas based on the signals Jay picks up in the opening conversation. A first-time renter looks different to a downsizing retiree. A bereavement triggers different needs to a first-time purchase.

“We’ve built a lot of empathy into the model because not every move is a happy move,” Ross said. “You might be getting divorced or there could be a death, just lots of different reasons, and the AI can pick up on that.”

The checklist Jay builds for each customer is dynamic – tailored to who they are, not just what they’ve asked for. Whether the customer stays with Jay or moves to a human specialist, nobody starts from scratch.

Energy runs as a co-pilot

“You have a co-pilot flow where you’re coming to the conversation, Jay’s making recommendations about what you should be doing next,” Ross said.

Jay handles the change of occupancy with the existing supplier, then recommends a new one. At each stage, the customer sees options alongside a chat interface and decides what to act on. For straightforward questions – explaining council tax to an international mover, notifying a supplier – Jay handles those directly. This is based on energy, the first live vertical – still early in rollout.

Just Move In sells regulated products – they’re directly authorised by the FCA to sell insurance.

“They’re happy to be served by AI up until a point,” Ross said. “But there will be this threshold where if it’s a particular question about a medical condition or something that is expensive, you still might want to speak to a human to make sure things are okay.”

Where the human team fits now

By the time a specialist picks up, the customer has already explained their situation and seen their options. “They’ll manage more customers in a day, but actually the interactions will just be a bit more condensed,” Ross said.

Jay is also booking appointments on its own, including out of hours. The conversion rates on those bookings are similar to what the human team achieves. Early signs suggest the customer experience hasn’t suffered – the Trust Pilot score has held at 4.9 through the transition.

How the review loop catches mistakes before the next vertical

The team is still in the early stages. Every conversation gets reviewed through a dual loop. The AI evaluates its own interactions, and humans evaluate them too. When something goes wrong, the correction feeds back into the model.

“We can see where the AI gets it wrong, then we can say to the AI, you made a mistake there. This is what you should do next time,” Ross said.

“And of course it was breaking a lot,” Ross admitted. “There are going to be a lot of bugs when these new tools roll out because they don’t exist. We’re something that is completely new.”

Just Move In are learning from energy before they switch on the next vertical. Rather than bolting on a feedback loop after deployment, they built it alongside the product.